Pendekatan Data Driven dalam Model Regresi Pengaruh Penggunaan Chatgpt terhadap Emosional Gen-Z

Abstract

Data were collected through a questionnaire from 500 respondents, using the DASS-21 instrument to measure mental health, as well as questions related to the frequency, duration, and purpose of ChatGPT use (schoolwork, entertainment/personal). Statistical analysis included descriptive analysis, instrument validity and reliability tests (Cronbach's Alpha > 0.900), and multiple linear regression. ChatGPT use for schoolwork was consistently positively correlated with depression, anxiety, and stress. Conversely, frequency of use in the past 7 days and use for entertainment/personal tended to be negatively or not significantly related to DASS-21 scores, except for entertainment/personal use which was significantly positive for depression. Duration of use per day was significantly positively related to depression and stress, but not significantly for anxiety. Academic use has the potential to increase the risk of depression, anxiety, and stress, while more frequent use in general or for entertainment/personal may have different effects. The results showed that ChatGPT use variables significantly predicted levels of depression, anxiety, and stress (R-squared 0.049-0.200). This research highlights the complexity of the impact of generative AI on Gen-Z's mental health and emphasizes the need for responsible use of AI.